Research On The Measurement Method Of Air Floatation Position Based On Computer Vision

In the process of ground simulation of the spacecraft,it is necessary to use an air floating platform to provide a simulation experiment environment.In the process of using the air floating platform for spacecraft dynamics simulation,it is necessary to obtain the position and attitude information of the air floating platform in real time to complete the control of the position and attitude of the air floating platform.Based on this research background,this subject uses computer vision related measurement principles to study the position and attitude measurement methods of the air floatation platform,and provides a non-contact,low-cost,high-precision air floatation platform position and attitude measurement method.First,the camera modeling and calibration methods are studied.The imaging principles of the monocular camera and binocular camera are analyzed,and the pinhole imaging model,lens distortion model,triangulation ranging model and stereo correction model of the camera are established.The corresponding calibration algorithm is given,and the unknown parameters in the model are obtained through the camera calibration experiment.Then,the method of target feature extraction and matching is studied.For cooperative targets,the sub-pixel centroid positioning method based on two-dimensional Gaussian surface fitting and the landmark point matching algorithm based on a priori geometric relationships are studied;for non-cooperative targets,the feature point extraction algorithm based on SIFT feature descriptors And feature point matching algorithm based on the principle of random sampling consistency.After that,the method of posture measurement for cooperative targets is studied.A method for converting measurement data from a two-dimensional image coordinate system to a three-dimensional world coordinate system is given.A posture measurement model based on cooperative target is established,and a posture measurement method based on iterative update is proposed.Aiming at the problem of precision calibration of the cooperative target posture measurement system,a precision calibration device and method of the posture measurement system are proposed,and the precision of the system's posture measurement accuracy can be calibrated only by collecting test images once.According to the proposed pose measurement method and accuracy calibration method,the verification experiment of the position measurement accuracy and attitude measurement accuracy of the cooperative target is completed,and the measurement accuracy of the algorithm is verified.Finally,the method of pose measurement for non-cooperative targets is studied.A method for converting measurement data from a two-dimensional image coordinate system to a three-dimensional world coordinate system from a binocular measurement system is given.A pose measurement model based on non-cooperative targets is established,and a pose measurement method for non-cooperative targets based on least squares fitting error and SVD decomposition is given.Aiming at the accuracy calibration problem of non-cooperative target pose measurement system,a calibration method for pose accuracy based on Gazebo simulation environment is proposed.According to the proposed pose measurement method and precision calibration method,the verification experiment of the position measurement accuracy and attitude measurement accuracy of the non-cooperative target is completed,and the measurement accuracy of the algorithm is verified.